Curing profile optimization of silver epoxy die attach in ball grid array package process / Ng Qian Qing

Ng , Qian Qing (2024) Curing profile optimization of silver epoxy die attach in ball grid array package process / Ng Qian Qing. Masters thesis, Universiti Malaya.

	PDF (The Candidate's Agreement) Restricted to Repository staff only Download (169Kb)
	PDF (Thesis M.A.) Download (5Mb)

Abstract

Ball grid array (BGA) packages are essential components in automotive applications, renowned for their compact form factor and high-density interconnections. However, the persistent issue of void formation within these packages poses a significant challenge to their performance and reliability. This challenge is primarily rooted in incomplete reactions between the die attach epoxy and the resin, resulting in the presence of unreacted curing agents. When exposed to moisture, these residual agents can trigger substrate surface degradation, potentially leading to critical electrical failures. This research endeavours to address the challenge of void reduction in 3‐IC‐Chip‐ MAPBGA packages, a subset of Ball‐Grid‐Array (BGA) packages, through systematic manipulation of epoxy curing parameters. The silver conductive epoxy used in this study possesses a composite structure consisting of cross-linked polymers with strong covalent bonds. Six cells, each containing eight samples, underwent testing. Four of these cells were subjected to two-step curing profiles, while two cells, including a control sample, were treated with single-step profiles. Post‐curing, comprehensive performance testing revealed the superior efficacy of two‐step profiles in void mitigation. The two-step curing profile is better than the single step curing profile because it allowed for a controlled, gradual curing process that reduced internal stress, improved outgassing of volatile components, and ensured thorough, uniform cross-linking of the epoxy. This resulted in fewer voids and enhanced the overall reliability and performance of the BGA packages. Specifically, Cell#3 and Cell#4, both utilizing two-step curing profiles with different first-step temperatures, exhibited minimal void formation percentages of 4.22% and 3.61%, respectively, compared to the control sample's 10.33%. The highest shear strength, reaching 97 MPa, was observed for Cell#4 at 25 °C. Additionally, Cell#4 exhibited a minimal percentage degradation of 3.6% in die shear strength when tested at the elevated temperature of 80 °C. The processing time for Cell#4 was found to be 105 min, the second shortest after the control sample, which had a processing time of 60 min. Overall, Cell#4, processed through a two‐step curing method, demonstrated optimal performance. Its curing conditions with higher first step ramp time and lower first step temperature is highly recommended for BGA packages, especially in high‐density interconnect applications within compact ICs, a common feature in the automobile industry. These findings pinpoint the optimal curing profile for the die attach process, strategically aimed at minimizing void formation within the package.

Actions (For repository staff only : Login required)